OMRI BOEHM | Kant’s Critique of Spinoza | Oxford University Press USA 2014


By Sebastian Gardner 

Omri Boehm offers a lucid and incisive defence, supported by careful scholarship, of the compelling idea that preoccupation with Spinoza—appreciation of the force of Spinoza’s reasoning along with a concern to avoid his drastic conclusions—is at the heart of Kant’s philosophical enterprise. I have learned a great deal from Boehm’s fascinating study, and its excellence will be clearly visible to anyone who has pursued the question of what the Critique of Pure Reason is aiming to achieve. Since a full engagement with all of the themes treated in the book is out of the question, I am going to focus (in Sections 1 and 2) on Boehm’s treatment of the pre-Critical work in which, he argues, Spinozism first becomes a major issue for Kant, and then (in Section 3) say a few things about Boehm’s general account of Kant’s confrontation with Spinoza and indicate how his configuration of the Kant-Spinoza opposition sets the stage for a later chapter in the development of classical German philosophy.

It will be helpful if I outline my differences from Boehm at the outset. Boehm launches his enquiry by arguing that the Principle of Sufficient Reason (PSR) plays a key role in the argument of Kant’s early The Only Possible Argument in Support of a Demonstration of the Existence of God (hereafter, Beweisgrund). Below I explain why I am not persuaded that this is the case and why it does not seem to me that the core argument of that work—Kant’s remarkable attempt in Section One of the Beweisgrund to prove a priori the existence of a necessary being without assuming existence to be a predicate—renders Spinozism strictly unavoidable. This makes some difference to Boehm’s historical narrative, insofar as Boehm presents Kant’s Critical turn as a reaction against his own early rehearsal of Spinozism but it leaves intact Boehm’s broader thesis that Spinoza is constantly in Kant’s sights and central to the motivation for transcendental idealism, explicit reference to Spinoza being no measure of his importance for Kant.

Nonetheless one may have doubts about the extent to which Boehm allows Spinoza to dominate the narrative of Kant. Is it necessary to allow as much as Boehm suggests to rest squarely on the motive of resisting Spinozism? The tendency, which was especially noticeable in earlier anglophone Kant scholarship, to lopsidedly read Kant as occupied merely with the shortfall of empiricism and the problems posed by Cartesian and empiricist scepticism, certainly stood in need of correction, which recent scholarship has done a lot to provide, and Boehm’s book is a major further contribution to a deepening appreciation of the importance of Kant’s relation to early modern rationalism. But we do not lose anything, as far as I can see, by viewing Kant as engaged in a complex multi-sided manoeuvre, whereby the deficiencies of several philosophical positions are resolved at a single stroke—the greater the number of early modern philosophers whose difficulties can be exposed as deriving from their tacit commitment to transcendental realism, the stronger the case for transcendental idealism. Reading the Critique of Pure Reason as a critique of Spinoza should not lead us to forget that it is also and equally a critique of Hume and Leibniz, among others.[1] Kant’s eggs are not all in one basket. Defeating Spinozism may be the ultimate objective, which would crown the achievement of the First Critique, but its philosophical success does not rest on that claim alone: it would be enough if its upshot were simply a refinement of the alternatives—which would accord well with the historical record, insofar as the post-Kantian development involves a fusion of Kant and Spinoza.

1. Kant’s Beweisgrund and the Principle of Sufficient Reason

Boehm interprets Section One of the Beweisgrund as revealing an ab initio commitment to the PSR. It will help to have Boehm’s reconstruction of the argument of the Second and Third Reflections before our eyes (pp. 20–1):

1. Internal possibility (the essence of a thing) depends on formal and material possibility.

2. Formal possibility (the logical consistency between a concept’s predicates) depends on material possibility (the predicates themselves).

From these definitional steps Kant continues to elaborate his argument:

3. Material possibility is grounded in something actually existing.

4. Necessarily, something is possible.

5. Necessarily, something exists. [From 3 and 4]

6. There is a being that exists necessarily.

7. There can be only one necessary being.

Boehm maintains that in order to complete this argument Kant draws tacitly on the PSR (pp. 16, 21–9). The importance of this claim is of course that, if correct, it shows Kant to have put himself under pressure to complete the (arguably) inexorable movement compelled by the PSR to Spinoza’s necessitarian substance monism. Appeal to the PSR is required by Kant, according to Boehm, at four crucial points:

(i) First, in establishing the dependence of possibility on actual being (D3). Boehm describes D3 as “plausible” in the light of D2 but adds that it “also relies on the PSR” (p. 24). In explaining how and why the PSR comes in, Boehm looks ahead to the First Critique and cites Kant’s statement at A308/B364 that if the conditioned is given, then so too is the series of conditions, hence the unconditioned, and applies this PSR-like principle to the case at hand: something must actually exist, “[o]therwise the fact that something is possible will not be ultimately explained, which is rejected by the PSR” (p. 25). The argument is later reiterated:

[B]y the PSR, all possibilities must be grounded [in being, the context makes clear]. For if some possibilities weren’t grounded [in being], there would be inexplicable possibilities, which is rejected by the PSR. (p. 28)

Arguably the requirement that modal facts supervene on existents is a separate principle, independent from the PSR. If the PSR is stated simply in the standard form of the requirement that nothing be (allowed to be thought to be) the case without (its being thought that there is) sufficient reason for its being the case, i.e., simply as answering ‘why?’ questions, then it does not of itself tell us what ontological status, if any, is to be assigned to whatever it is that counts as sufficient reason. In other words, PSR in its raw unelaborated form does not tell us what is required to qualify as a Grund and so its application to possibility does not tell us whether or not possibility has its sufficient Grund in the existence of any being. Wherever our knowledge that the material component of possibility must be “given as existing” (p. 24) might come from, it cannot be from the PSR alone. The groundedness of possibility in actual being (D3) is logically presupposed by any ontologically significant employment of the PSR, not derived from it.[2]

(ii) Second, regarding the necessity that something be possible (D4), Boehm observes that the argument which Kant gives for this proposition may seem a “trick of words” and goes on to explain how his reasoning “can also be supported by the PSR”, even though Kant does not appeal to it (pp. 26–7). Boehm reconstructs Kant’s argument as follows: (1) The PSR requires that “modal claims be fully explained”. (2) Absolute impossibility, if it were a modal fact, could not be explained. Whence (3) the impossibility of absolute impossibility (D4).

An initial problem here is that it is not clear that something of the order of absolute impossibility could be required properly to submit itself to the PSR for “explanation”, in other words, that a proposition which defines the limiting framework of modality can be taken as asserting a “state of affairs” (p. 26). But even if it is granted that absolute impossibility would constitute an explanation-worthy state of affairs demanding application of the PSR, it may still be held that it could satisfy the PSR in a direct reflexive manner: if absolute impossibility obtained, then the pure modal fact that nothing is possible, it may be said, would sufficiently explain why nothing is possible. To resume the earlier point, if the notion of absolute impossibility is (so to speak) absolutely repugnant, then the proper conclusion to draw may be simply that we are face to face here with the sheer ineliminability of possibility, in parallel with the sheer impossibility of the truth of a formal contradiction. If so, the ineliminability of possibility cannot be regarded as an explanandum to which the PSR may be applied, any more than the PSR can be applied, presumably, to the necessity that there be sufficient reason for all that is really, extra-logically, the case.[3] To call the ineliminability of possibility a ‘brute fact’ of a sort that the PSR precludes would be to overshoot the mark, since if the putative ineliminability of possibility counts as a brute fact, then so too does the putative fact of the necessity of the conformity of all real things to PSR, in which case the PSR would have to be declared contradictory and self-refuting.

Kant’s own presentation of the case for D4, I think, suffices as it stands.[4] Kant’s claim, as I understand it, is that thinking manifests immediately the reality of possibility in a way similar to that in which, according to Descartes, it manifests the reality of a thinker, that is, without any inference (or at least, without inference from one existent to another). If the reality of possibility is testified directly by our thinking, then no principle of thought, such as the PSR, is needed to rule out absolute impossibility. If so, Kant’s claim in D4 is not, as Boehm puts it, that it is analytically false or “inconsistent” to say that absolutely nothing is possible (p. 26), rather absolute impossibility is excluded before we get to the point of being able to determine relations of logical (in)compatibility.

(iii) The PSR is required next, according to Boehm (pp. 27–32), in order to move from the necessity that something should exist (D5) to the existence of something that exists necessarily (D6). Again Boehm notes that the inference may be contested, and after rejecting on textual grounds Adams’s reconstruction of Kant’s argument, he proceeds to offer a defence of it in terms of the PSR, while noting that Kant himself does not even attempt to justify it (pp. 28–9).

Again it seems to me that the PSR is not required for the inference. It follows already, from the earlier conclusion that possibility enjoys non-contingent reality, that whatever being makes possibility possible (D5) must be considered, by virtue of its occupying that role, to exist necessarily (D6). Possibility-grounding actual being cannot be thought to exist non-necessarily, for if it existed contingently, then it would be possible to remove it in thought, to think it away, which would be for thought to cancel its own possibility. If the being which subvenes possibility is irremovable—if its non-existence is unthinkable—then it must be thought to exist necessarily. The inference is open to challenge insofar as it involves a movement from a necessity pertaining to thought to a necessity pertaining to what thought identifies as its ground, but it is neither obviously valid nor obviously invalid, and I suggest that it must remain in this condition of undecidedness for as long as our general understanding of the relation between those two species of necessity, which can be neither collapsed into one another nor absolutely dissociated, remains incomplete.

(iv) As regards the final move which Kant’s argument requires, to the singleness or uniqueness of the being which exists necessarily (D7), Boehm again makes this a matter of the explanatory requirements of the PSR, concerning, he argues, the need for the interrelations of all possibilities to possess a sufficient ground (p. 28).

Once again it seems to me that invocation of the PSR, which Kant gives no sign of relying on, does not help him. Kant’s brief and dense argument in Section 3 of the Third Reflection (BDG, AA 2:84–5) is obscure but does at least make it clear that it is based on the preceding discussion of the concept of absolutely necessary being, the baseline of which is the provision of das Materiale or Data for thought (BDG, AA 2:81–3). The following reconstruction, according to which Kant’s argument turns on the interconnection of the concept of absolutely necessary being with that of a source of possibility, goes beyond Kant’s text but makes sense of his reasoning. If necessary being were plural, then the question would arise, regarding each Necessary Being, whether it is sufficient for the reality of possibility. If each were sufficient, then each individually could be thought to be unnecessary for the reality of possibility and so could be removed in thought, since the other Necessary Beings that remain could be thought to bear the load. But if each in turn could be removed qua source of possibility, then sequentially all could be removed, which contradicts D4. If the hypothesis of multiple Necessary Beings is to avoid this consequence, then this must be either (a) because their sufficiency for possibility results from their conjunction, in which case they compose a single Necessary Being of which they are parts or aspects, or (b) because there is some one Necessary Being which ensures that thought is furnished with its data, in which case it is this guarantor of possibility which qualifies as the single irremovable nothwendiges Wesen whose existence is asserted in the final step of the argument (D7), and whatever other Necessary Beings may be in the vicinity fall away, as lying outside the scope of the proof. The hypothesis of a plurality of necessary beings thus resolves itself into the hypothesis of a single Necessary Being, and once again the motor of Kant’s argument is his thesis that there must be possibility and that it must derive from actual being. As before, it may be objected, here in Third Paralogism style, that the argument involves an illicit movement from singleness qua source of possibility to singleness per se, but if this does render the inference problematic, then the PSR will not restore its validity, because the PSR does not of itself tell us how the sufficiency of grounds is converted into the individuation of entities.

As I have taken pains to indicate, the reconstructions that I have offered of Kant’s inferences may be challenged as regards their validity, all roughly in the same way, but they are not, I maintain, non-starters, and if I am right that Boehm’s application of the PSR does not yield better results, then their inconclusiveness does not count against them in the present context. What should also be stressed is that the primary question we are concerned with here, recall, is not what argument might best deliver Kant’s conclusions but how, from the evidence, Kant intended to argue, and what the text suggests strongly is that the nub of the Third Reflection is supposed to be contained in the analysis of possibility in the Second Reflection. As I read Kant, he means to argue directly from (1) the account of possibility as having a material as well as a formal component, which must be given to thought, and given as existing, and from (2) the necessity of possibility which is implied immediately by mere thinking, to (3) the existence of a Necessary Being; where the new principle driving his proof operates along the dimension, not of relations between thoughts or elements within them, as do the PSR and the Principle of Contradiction, but of (compatibility with and grounding of) the possibility of (its being true that) anything is being thought, or that thinking can take place, at all. It is surely of high relevance that even the necessary falsity of a contradiction (hence formal possibility too) is resolved by Kant into the exact same root.[5]

Now this construal of the argument of Section One of the Beweisgrund will ring loud bells, since its fulcrum lies in consideration of what makes determinate thinking possible, where this refers to a type of grounding which is neither logical in the narrow sense (formal logic establishes only ‘formal possibility’) nor a matter of worldly causality—and exactly this is also of course the linchpin of what Kant later calls transcendental proof, which operates on the basis of sheer possibility (now that of Erfahrung) and issues in synthetic a priori propositions. At a finer level of detail there is a parallel to be drawn between the impossibility of thinking away the material conditions for thought asserted in the Beweisgrund, and the irremovability and consequent necessity of space and time asserted in their metaphysical expositions in the Transcendental Aesthetic of the First Critique.[6] What is most striking in the Beweisgrund is the obliqueness of the (elusive) necessity in thought that Kant wants to put to work: Kant’s idea is not that we cannot think away our own existence or that of our thoughts—there is no necessity in the existence of either of those objects—but that we cannot think away the situation of thinking’s being possible.[7] This pure structure of thinkability is internal to thought and imposes itself on us with a distinctive type of necessity, which in the Critique of Pure Reason Kant will try to account for immanently, but which in the Beweisgrund is taken to reveal the immediate anchoring of thought in reality, as no mere consideration of the agreement of concepts can do, and yet in a way which involves none of the mediation required for a posteriori cognition. This profoundly original element does not come to light when the text is read through Spinoza’s eyes.[8]

The notion that the Beweisgrund is genuinely distinguished from early modern rationalism with regard to its method would seem to be testified by Jacobi’s ecstatic reception of the work recounted in his David Hume (1787) (Jacobi 1994:284–5), and its inspiration of the seminal ideas sketched in Herder’s ‘Versuch über das Sein’ (1763).[9] That Kant would have had a very strong interest in uncovering a new a priori epistemic source independent of the PSR (and of the Principles of Identity and Non-Contradiction) coheres with the misgivings about Wolff’s use of the PSR that he had expressed in the New Elucidation (1755),[10] and also with the fact that, as L.W. Beck (1969:409–10) tells us, the notion that possibility presupposes actuality “had become almost a commonplace” by 1763. It would be puzzling if Kant intended the Beweisgrund to do no more than merely rework a received idea.[11] If on the other hand the Beweisgrund embodies what Kant considers to be a radical insight at a foundational level—with justification, insofar as we can see it to contain the seed of the Transcendental Turn—then its ambition is explained. The argumentative shortfall of the Beweisgrund, on the view I am offering, reflects no simple fallacy in Kant’s reasoning but derives from uncertainty at the root of the argument concerning what exactly it means for thought to recognise that there is something which it must conceive as having absolute but not logically necessary existence. It is reasonable to suppose that Kant’s awareness of this limitation, which makes the argument of the Beweisgrund inconclusive—but which, to repeat, does not mean that it is based on an outright mistake, hence does not require that it be discarded—added impetus to the formation of the Critical concept of a transcendental ground.

2. Being and Possibility in the Beweisgrund

Disputing Boehm’s account of how Kant arrives at his conclusions in the Beweisgrund does not mean denying the work Spinozistic significance. The Beweisgrund creates space for Spinozism, I suggest, not because of any implicit appeal to the PSR, but insofar as Kant’s affirmation of the absolute priority of being over possibility raises acutely the difficult question of what possibility is, of how it gets into the picture seemingly in addition to being and its determinations: a question which lays the ground for the Spinozist to argue that we should deny it any such additional reality. Only in this weak and indirect sense, I think, can its implicit logic be described as Spinozistic. That there is this potential for Spinozism in the Beweisgrund does not mean however that it was actualised. Two alternative views of the Second and Third Reflections, one weaker and one stronger, seem to me possible and textually warranted.

On the weaker view, the Third Reflection leaves undecided how exactly the grounding relation of being and possibility should be conceived. It is surely important that Kant disavows any claim to be offering in the Beweisgrund a complete metaphysical position which would answer all of our questions concerning Reality; the aim is only (!) to establish a proof of the existence of God. Second, Kant does not of course think that the argument of the Third Reflection of the Beweisgrund suffices on its own as a theistic proof: it merely lays the ground for further work, which Kant executes in the Fourth Reflection of Section One, where he argues that the Necessary Being must have properties of understanding and will, and in Section Two, which re-argues this claim on an a posteriori basis. These additionally imputed properties involve realism regarding possibility of a sort that the Spinozist rejects, since if the Necessary Being were to exhaust Reality überhaupt and coincide with it in all thinkable respects, then nothing would remain which could provide logical opportunity for an exercise of rational volition displaying goodness, and ascription of that power to God would be without meaning.

Clearly, if this is Kant’s strategy, then Kant must suppose that his proof of a Necessary Being does not preclude the realism about possibility required by theism. It may be asked how Kant can start on the business of showing that the Necessary Being of the Third Reflection has additional, theistic properties without having already established the realism about possibility which they presuppose. My answer is that Kant rightly thinks that he does not need to come up with a refutation of the Spinozist’s modal anti-realism, but only to have left the matter open at the end of the Third Reflection—that is, he assumes there is no manifestly compelling argument for anti-realism regarding possibility. We can suppose moreover Kant to be aware that, if he is wrong about this, then indeed a proof of the existence of any God other than Spinoza’s is impossible. But again this would mean not that the primary task for any proof of the existence of God is to refute the Spinozist’s modal anti-realism, but that the onus lies on the Spinozist to establish it. What makes this supposition reasonable is that it is far from obvious that possibility cannot amount to anything more than the inherence of properties in existents: the anti-realist view of possibility is highly revisionary—it requires nothing less than the whole complex machinery of at the very least Book One of the Ethics—and cannot be taken off the peg. If on the other hand Kant’s reasoning in the Third Reflection follows the course described by Boehm,[12] then the argument of the Beweisgrund incorporates an error of considerable magnitude which it is hard to account for.[13]

On the stronger view of the being-possibility relation, Kant’s position in the Beweisgrund incorporates a positive non-Spinozistic commitment. Boehm maintains that Kant is obliged to identify possibility in general outright with determinations of the Necessary Being, that is, with “properties” that “inhere” in it (pp. 25, 30–2, 38). Again Boehm grants that this is not a textually explicit part of Kant’s reasoning but an interpolation, to be justified on the basis that the only alternative is indefensible. Yet as Boehm notes (p. 25), Kant sketches not one but two construals of the grounding of possibility:

The data of all possibility must be found in the necessary being either as determinations of it, or as consequences which are given through the necessary being as the ultimate real ground. It is thus apparent that all reality is, in one way or another, embraced by the ultimate real ground.

Da die Data zu aller Möglichkeiten in ihm anzutreffen sein müssen, entweder als Bestimmungen desselben, oder als Folgen, die durch ihn als den ersten Realgrund gegeben sind, so sieht man, daß alle Realität auf eine oder andere Art durch ihn begriffen sind. (BDG, AA 2:85; emphasis added)[14]

Boehm argues that only the first construal makes sense, and to assert that possibilities are grounded on Bestimmungen, according to Boehm, is to say that they just are determinations or properties of the Necessary Being: “possibilities grounded in determinations are thus possibilities because they inhere in the existing being” (p. 30; emphasis added). To instead ground possibility on the “consequences” (Folgen) of the Necessary Being fails to make sense, according to Boehm, because (1) the Folgen in question would perforce be “finite beings”, “ontologically separate” from the Necessary Being, such that (2) the net result would be to reduce Kant’s proof of God in the Beweisgrund to the familiar cosmological argument, which cannot have been Kant’s intention (p. 31).

It is not immediately obvious what licenses the transition from “possibilities are grounded in determinations”, which is what Kant says, to “possibilities are determinations”, nor why allowing Folgen to mediate the relation of possibilities collapses the Beweisgrund into the old cosmological argument.[15] Why cannot Kant’s positing of Folgen simply be an elaboration of the conclusion, already reached, that possibility is grounded in a Necessary Being? Boehm puts a lot of weight on the allegedly problematic character of ontological separateness, and shortly I shall say why I think this is unjustified, but I concede that the Beweisgrund is “somewhat unclear” about the grounding relation (p. 30). Boehm’s Spinozistic clarification of Kant seems however to eliminate too much. Though aware that this type of relation is less transparent than simple inherence (BDG, AA 2:80), what Kant wants is without question—or so it seems to me—some kind of supervenience: Kant goes out of his way to show that there must be slack in the relation of possibilities to their ground, and immediately after the passage quoted above spends two pages explaining that what does not follow from his earlier argument is Boehm’s Spinozistic reduction:

But this is not to be understood to mean that all possible reality is included among its determinations [so ist dieses nicht so zu verstehen, daß all mögliche Realität zu seinen Bestimmungen gehöre]. This is a conceptual confusion [eine Vermengung der Begriffe] which has been uncommonly prevalent until now. (BDG, AA 2:85; emphasis added)

As Kant argues the point: Provision must be made for “real opposition” in what we find to be the case—e.g., opposing forces in a physical body, or the sensation of pain—and also for negations and defects—e.g., lack of the power of thought—for these are among the things whose possibility the Necessary Being must provide for. This can be done only if we avoid taking such items as indices of, i.e., as licensing inference to, either (a) “logical contradictions”, i.e., contradictory predicates, within the ground of possibility, or (b) “real opposition or positive conflict among its determinations”, since this would signal defectiveness in the Necessary Being (BDG, AA 2:86). What follows according to Kant is that (i) not all “possible reality” is “included among” the determinations of the Necessary Being, (ii) certain realities do not exist in the Necessary Being as determinations thereof (“so können sie nicht insgesammt als Prädicate in ihm sein”), and (iii) certain negative states of affairs or defects “depend upon” and are “grounded in” the Necessary Being, with respect to “what is real in them”, yet are not predicable of it (BDG, AA 2:86–7). The key notion therefore is that these other realities are given through the Necessary Being (“weil sie doch alle durch ihn gegeben sind”), and as such belong to it in some either direct or indirect sense (“so werden sie entweder zu seinen Bestimmungen oder Folgen gehören”), but are not given as in it in the same manner as its properties.[16]

That Kant would find it natural to posit an intermediate realm between God as the Necessary Being, and the finite worldly objects of our immediate cognition, is surely plausible in light of the Leibnizian-Wolffian background.[17] As Boehm notes (p. 35), the primary example Kant gives of something which must be assumed in order to provide thought with its data but whose possibility cannot be further analysed, is extension (BDG, AA 2:80–1). In the language of Wolff and Baumgarten, and contemporaries of Kant’s such as Lambert and Mendelssohn, extension is a Realität, a notion employed freely in the Beweisgrund.[18] The sense in which Realitäten such as extension, which cannot be ascribed to God, are consequently ‘ontologically separate’ is a relatively fine matter: they are at any rate not separate in the manner of objects, such as material bodies, that can exhibit real opposition. From the evidence, as far as I can see, Kant in 1763 had no compelling basis for thinking Realitäten incoherent and did not feel the pressure of the issue of ontological separateness highlighted by Boehm, decisive though it may be for the argument between Leibniz and Spinoza.[19]

One important consideration advanced by Boehm is that Kant later describes the implications of the argument of Reflection Three of the Beweisgrund as Spinozistic. I suggest this has a straightforward interpretation consistent with what I have proposed. On the weaker of the two views I described, Kant has no full positive theory in the Beweisgrund of how possibility can supervene on, without being reducible to, actual being, and Boehm is exactly right that it is here that the Spinozist does well to apply pressure. Nevertheless and not unreasonably, I argued, Kant in the Beweisgrund supposes that the picture that he requires is intelligible and well motivated. This is what changes in the Critical period, where Kant has given up on the idea that gaps in ‘dogmatic’ metaphysics can be filled by positive theorising, and has transferred its burden to the practical domain, so there is no longer reason to maintain the thinkability of alternatives to Spinozism. And canvassing Spinozism as the sole alternative to transcendental idealism—raising the stakes and writing the Leibnizian-Wolffian compromise, as Kant wants it to appear after 1781, out of the picture—makes the Kantian option harder to decline: which is Kant’s explicit strategy in the Second Critique. The stronger view of the possibility-being relation can therefore also be abandoned in the Critical philosophy.

A similar kind of explanation can be given for the fact that Kant recapitulates the reasoning of the Second and Third Reflections of Section One of the Beweisgrund in the Transcendental Ideal of the First Critique. It is essential for Kant’s general strategy in dealing with transcendent metaphysics that an Idea of Reason be located at the root of the concept of God, in order for it not to be an arbitrary construction, and the line of reasoning which he had developed earlier in the Beweisgrund—and which on my reading rests on extrapolation from the fact of finding oneself in the bare situation of being a thinker of would-be determinate truths—supplies Kant with what he needs: Kant is able to claim that we are led from determinate thought to the Idea of an ens realissimum, in the same way that we are led from experience in general to the Idea of a transcendentally free cause. Just as Boehm reads the Beweisgrund as actual Spinozism, while I read it as at most opening the door to it, Boehm sees the Transcendental Ideal as rehearsing Spinozistic reasoning specifically in order to disempower it, while I interpret Kant in this part of the Critique as revisiting his earlier argumentation, conserving the part of it that he continues to regard as cogent, viz., the Gedankengang that leads from thought to the Idea of a ground of all thinkable possibility, while denying its conclusion the ‘dogmatic’ status that he accorded it in 1763; such that, while of course the Transcendental Ideal in the full context of the Dialectic has critical force vis-à-vis Spinoza, this is only one aspect of it.

3. Kant vs./& Spinoza

The foregoing relates to Boehm’s claim that Spinozism is, so to speak, internal to Kant and by abreaction a driver of the Critical project. The discussion which occupies the other main chapters of Boehm’s book, and which largely stands independently of the claims of Chapter One, is devoted to the question of how the Critical Kant means to deal with Spinoza and with what degree of success.

Boehm refers to the illuminating discussion by Fisher and Watkins (1998) of the relation of the Beweisgrund to the Ideal of Pure Reason (p. 50). Fisher and Watkins emphasise the difficulty of locating Kant’s justification for denying constitutive status to the ens realissimum, and suggest that it has to do with Kant’s account of reason and the understanding, while making it clear that this does not by any means render Kant’s argument irresistible. Boehm’s own solution—namely, that the theory of transcendental illusion is what undercuts the security of the inference to D6, warranting the conversion of Spinozism into merely regulative doctrine—is similarly indirect and tentative: Boehm grants that it remains a “vexing question” whether the theory of transcendental illusion is sufficient to deflect Spinozism (p. 58) and accordingly moves the discussion on to the Antinomy of Pure Reason, where he considers that Kant’s more compelling criticisms of Spinoza are to be found.

A general question may be raised concerning Boehm’s strategy of allowing the Antinomy of Pure Reason to play the lead role in Kant’s endeavour to refute Spinoza. It is worth pointing out that in Kant’s head-on confrontation with Spinoza in the Second Critique (where, Kant tells us, we come to see the full importance of the critique of speculative reason), Kant refers to time’s ideality qua “mere form of sensible intuition”, i.e., to the grounds for idealism presented in the Transcendental Aesthetic, and not to the contradictions of rational cosmology (see KpV, AA 5:100–3). Kant’s accent here is not on any inherent defect in the notion that time is a fundamental determination of reality, but on the independence of the temporal way of representing things from whatever determinations they receive from the “original being”, and this is a positive claim about time which requires the positive epistemology of the Aesthetic, not a direct consequence of the negative argument of the Antinomy.

As regards the crunch question of how far Kant actually gets in his endeavour to undermine Spinoza, Boehm’s answer is affirmative: he concludes that the rationalist conviction that the world must be explicable through and through, which makes necessitarianism unavoidable and collapses Ought into Is, should yield to the anti-necessitarian demands of morality (specifically, of “moral outrage”), for the Kantian reason that Spinoza’s project has been shown to rest on a “normative decree rather than on theoretical justification”, a decree which is itself practically motivated and thus answerable to the demands of freedom (pp. 182–5). (This is a very bare statement of Boehm’s conclusion as it emerges from his subtle discussion of the ontological argument in Chapter Four. His full account involves also a highly interesting discussion in Chapter Two of the Spinozist’s right to the notion of an actual infinite, in which the sublime makes an unexpected appearance.)

Here however is one ground for worry. When Kant explains the value of the theory of non-temporal noumenal agency in the passage from the Second Critique referred to above, his claim is that we can we lay claim to real authorship of (and control over) our actions only if we suppose the temporal realm to be excluded from the scope of divine determination (God’s creation). The transcendental ideality of time is thus what grants us the ontological separateness, the scope to posit an Ought independently from an Is, that Spinoza denies us, and which, Kant affirms, Spinoza rightly regards as impossible in a realm of things in themselves, which we are obliged to conceive as deriving from and as fully determined by a single original being. The result of this manoeuvre is however, as several of Kant’s contemporaries quickly noted, highly paradoxical: Kant de-realises our agency as we sensibly cognise it, in order to relocate it, or rather its true ground, in the noumenal sphere, where, Kant asserts, absolute determination prevails. If this is right, Kant appears to have analysed the transcendental freedom required by pure practical reason into a mere appearance of contingency behind which lies a reality of intelligible fatalism. Either that, or our actions qua sensible are God-independent—whence the possibility of our true authorship—but not therefore ‘mere appearance’ in contrast with (a substrate of) things in themselves, and transcendental idealism is not the doctrine that it seemed to be; it now appears unnervingly close to what Fichte will shortly declare its real meaning to consist in.

Whether this affects Boehm’s account depends upon how much of a positive constructive alternative, on his view, Kant needs to come up with in order to counter Spinozism. But some challenge has emerged. The expectation was that Kant’s repudiation of the necessitarian PSR would be accompanied by a coherent alternative theoretical picture, but if the worry just expressed is correct, this is doubtful. The Spinozist would do well at this point to refer back to the Beweisgrund and ask how matters are supposed to have been improved. In Section Two of that work Kant attempted, hastily and rather desperately, to make provision for “actions which issue from freedom”, by postulating “an inadequately understood contingency” and “an indeterminacy in respect of determining grounds” (BDG, AA 2:110). The Spinozist may reasonably ask whether the Critical philosophy offers anything superior which would begin to cover the costs of renouncing the PSR.

If reclassifying the results of the exercise of pure reason as regulative does not strip them of their ‘dogmatic’ ontological force, and if transcendental idealism does not reconcile theoretical and practical reason, then more must be going on in pure reason than Kant has yet brought to light and a new attempt to reconcile freedom and reason must be made. Following this lead, let me finish by jumping ahead many decades. Boehm’s book is of course not a study of the post-Kantian development, and in claiming Schelling’s relevance here I just mean to indicate how the Kant-Spinoza opposition to which Boehm gives sharp definition supplies, historically and systematically, a basis for further philosophical development.

There is extraordinary continuity of Kant’s 1763 Beweisgrund with what is, chronologically speaking, the very last word in the development of classical German philosophy. Schelling is well known for the overtly Spinozistic system that he espoused between roughly 1801 and 1804, from which Hegel branches off, but his Philosophy of Identity, as he calls it, represents just one of Schelling’s several attempts to consummate his philosophical insights in a wissenschaftliche form, and if taken in isolation it occludes the profound problematic that really occupied Schelling from the beginning to the end of his philosophical career and which becomes fully clear only in his writings from the Weltalter period (c. 1811) onwards. This late philosophy can be viewed, I suggest, as an extended meditation on the insight that leads Kant to think that his modification of rationalism in the Beweisgrund opens a new door which puts the ultimately desired object of human reason within its grasp.

Schelling’s late view may be reconstructed as follows. There is a limit point at which philosophical reflection finds itself poised exactly between Spinoza and Kant, pointed simultaneously in both of their directions: that of supposing pure reason to have its own absolute content, reflective elaboration of which can encompass all actual and possible reality, inclusive of reflection itself; and that of acknowledging its own dependence on a Prius that it cannot retrieve or make transparent, a necessity which cannot be reduced to any logical relation. The former yields the necessitarian PSR of Spinoza, and returns (in a semi-degenerate form) in Hegel. The latter, guided by the intention of preserving contingency and freedom, is expressed in Kant’s assertion of the dependence of determinate truth on the independent contribution of sensible intuition—the equivalent in transcendental philosophy of the Data zu aller Möglichkeit in the Beweisgrund.

Now the problem, as Schelling sees it, is that neither can be repudiated. The necessity of the a priori has been demonstrated by the Critical philosophy, but Hegelianism exhibits the total emptiness that results ultimately from the endeavour to extract reality per se from pure reason. Recognition of the Hegelian terminus leaves us confronted with (in Schelling’s new idiom) the Fact of the World at the root of thought, and in effect takes us back to the situation of the pre-Critical Kant, caught between the claims of empiricism and rationalism (whose history Schelling’s 1830s texts go over repeatedly). Schelling is therefore recovering, and for the first time making adequately clear, the insight of the Beweisgrund that reflection on the possibility of thought leads to a point where we so to speak pass over into transcendental reality: If I am to think—and I do think!—then there must be possibility, and in order for there to be possibility, there must be some being which pre-dates thought and possibility, and hence which also pre-dates the PSR. The necessity which manifests itself at this point does not allow for a distinction to be drawn between the necessity that p (where modality is ascribed to the content of thought) and the necessity of thinking that p (where modality attaches to the thinking of the content). The PSR is thereby shown to be, in a newly disclosed sense, non-necessary, though it is also—and on the condition of being so understood—absolutely valid.

The highly complex structures that Schelling develops in his late philosophy are attempts to square Kant and Spinoza, not of course with respect to the letter of their writings, but with respect to their underlying philosophical dynamics. In what is virtually his last text, ‘Abhandlung über die Quelle der ewigen Wahrheiten’ (1850; see esp. SW, XI:585–90/1990:64–7), Schelling identifies the summit of previous reflection on the problem of necessary truth with Kant’s Ideal of Pure Reason, from careful consideration of which, he tries to show, the project of his own Positive Philosophy can be extracted. What Schelling might have added, if what I said earlier is correct, is that behind the Ideal lies the Beweisgrund, which stands on the cusp of grasping what Schelling calls unvordenklichen Seyn, insofar as Kant there recognises a necessity within thought which is more primitive than the PSR or Principles of Contradiction and Identity, but which, on Schelling’s diagnosis, a compelling (dialectical) illusion leads us to suppose manifests pure rational necessity. Schelling’s movement forward from Kant is to that extent also a return to an idea which Kant had glimpsed but left behind.

Invited: 24 October 2014; received: 10 August 2016.


[1] The point is made in Anthony Bruno’s review of Boehm, Notre Dame Philosophical Reviews, 11 November 2014.

[2] Relevant here is Schopenhauer’s treatment of the PSR in The Fourfold Root of the Principle of Sufficient Reason.

[3] Matters may be more complicated insofar as, arguably, reflexive applications of the PSR result in distinct, further modal truths and do not consist merely in different ways of stating the principle itself. But even granting this baroque ontology, it makes no difference to the point at hand, which is simply that, if the PSR can loop back on itself in a self-vindicatory manner, then so too can ‘Necessarily there is possibility’.

[4] More precisely, we could not be thinking in any sense that allows our thoughts to be candidates for truth: we could be ‘thinking’ only in the sense of executing purely formal, merely syntactic operations. In other words, Kant’s argument requires that our thinking be truth-directed (but not that any of our thoughts be true).

[5] “If we now consider for a moment why that which contradicts itself should be absolutely nothing and impossible, I find that through the cancellation [Aufhebung] of the law of contradiction, the ultimate logical ground of all that can be thought, all possibility vanishes, and there is nothing left to think [nichts dabei mehr zu denken sei]” (BDG, AA 2:82). In this paragraph Kant sets formal and material possibility in parallel and identifies the requirement that there be something zu Denken as the crux of both.

[6] Boehm notes the parallel at p. 61 n.18, but does not pursue it. Also worth noting as an anticipation of things to come is the argument at BDG AA 2:112, that “the commonly held concept of natural things, according to which their internal possibility is independent and without any external ground”—a.k.a. transcendental realism!—would make order in nature as good as miraculous, i.e., intrinsically unintelligible.

[7] Elucidation of these claims, and support for them, can be found in Kant’s discussion of the concept of existence in Section One, where Kant asserts both its unanalysability and its equivalence with the concept of “absolute position” or “absolute positing” (BDG, AA 2:70–6).

[8] The interpretation of the Beweisgrund sketched here differs also from those of other recent commentators, including Robert Adams, Andrew Chignell, Nicholas Stang and Peter Yong, but I shall not explore the differences here.

[9] The proto-transcendental character of the Beweisgrund is noted in Cassirer (1981:65).

[10] It is to be noted that at the end of the work, Kant distances himself from reliance on the PSR and also, significantly, endorses it only qua causal principle proceeding from an existentially committed premise (Section Three, BDG, AA 2:157–9).

[11] Or, for that matter, to merely repeat what he had said in Proposition VII of the New Elucidation. If we compare the two texts, the methodological advance is quite clear: the earlier text describes its proof as “based on essence” and as concerned with “the possibility itself of things”, not with the provision of data of thought (PND, AA 1:395).

[12] Boehm first says that Kant “was aware of” his commitment to substance monism (p. 7) but later qualifies the attribution (pp. 44–5) and finally suggests that “we should care less about Kant’s conscious commitments in 1763” (p. 45). This risks moving his account of the Beweisgrund out of the category of historical exegesis and into that of pure rational reconstruction.

[13] That Kant exhibits hesitation in this continuation of his argument, as Boehm rightly notes (pp. 42–3), is not probative: Kant may not think that the Necessary Being is ‘easily’ considered a person, but he cannot suppose the notion to have been ruled out.

[14] And note that “all reality is, in one way or another [auf eine oder andere Art], embraced by the ultimate real ground” (BDG, AA 2:87; emphasis added).

[15] Two further considerations advanced by Boehm in support of his account (p. 32) should be mentioned: Kant’s use of “enthält” to describe the relation of the Necessary Being to what it grounds, and his assertion in the First Critique that the ens realissimum comprehends (“in sich begreift”) all predicates not merely “under” but “within” itself (A577/B605). Regarding the first, it is notable that when Kant talks of enthalten, its object is, as far as I can see, not “all possibility”, as Boehm implies, but rather “den letzten Realgrund” thereof (BDG, AA 2:83, 84–5) or “die höchste Realität” (BDG, AA 2:85). Regarding the second point, it seems doubtful, as I say below, that the back-inference can be safely made, since in the Critical period Kant’s relation to Spinoza has changed.

[16] Understanding and will, by contrast, are true (wahre) realities (BDG, AA 2:87), and not Folgen but Bestimmungen of the Necessary Ground (BDG, AA 2:89).

[17] Indicative of Kant’s readiness to tolerate an intermediary in the grounding relation is his intermittent suggestion in the Third Reflection of a (conceptual or formal, but certainly not real) distinction between (a) the Necessary Being as such, and (b) “the ultimate real ground” (ersten Realgrund or letzten Realgrund) of possibility. As I emphasised earlier, Kant regards the latter as identical with the first—the Necessary Being must be considered “as” (als) the letzten Realgrund, for it is as such that we come to cognise it—but this does not make their concepts identical. Thus in the passage from BDG, AA 2:85 quoted above, Kant seems to be saying that what is “given as” (gegeben als) “the first real ground” (den ersten Realgrund) is not the Necessary Being itself but rather its (or perhaps just some of its) determinations and consequences. Also notable and of relevance here is Kant’s use of Realität as interchangeable with possibility in general at the beginning of the Fourth Reflection (BDG, AA 2:87).

[18] And that he had employed earlier in the New Elucidation, AA 1:395–6.

[19] Boehm writes: “Every contemporary of Kant’s would have to wonder how, or whether, Kant intends to evade the conclusion that extension just is a divine attribute” (p. 36). I agree that the question ought to have suggested itself but doubt that it did so in an urgent form, in view of the facility with which the notion of a Realität circulated. That said, the review which appeared in Briefe, die neueste Literatur betreffend in 1764 supports Boehm—but only insofar as the reviewer thinks that the a priori argument is pointless, since what does theistic work is the a posteriori argument of Section Two. That Kant in fact did not intend the Spinozistic conclusion, on the other hand, seems unquestionable, in view of Kant’s assertion that extension cannot be an attribute of that which has will and understanding (BDG, AA 2:85).


Beck, L.W. (1969), Early German Philosophy: Kant and His Predecessors (Bristol: Thoemmes Press).

Cassirer, E. (1981), Kant’s Life and Thought, trans. James Haden (New Haven: Yale University Press).

Fisher, M. and E. Watkins (1998), ‘Kant on the Material Ground of Possibility: From ”The Only Possible Argument” to the Critique of Pure Reason‘, Review of Metaphysics 52(2): 369–95.

Jacobi, F.H. (1994), The Main Philosophical Writings and the Novel ‘Allwill’, trans. and ed. G. di Giovanni (Montreal and Kingston: McGill-Queen’s University Press).

Schelling, F.W.J. (1856–61), ‘Abhandlung über die Quelle der ewigen Wahrheiten’, in Sämmtliche Werke (SW), ed. K. F. A. Schelling (Stuttgart: Cotta), vol. XI, pp. 573–90.

——— (1990), ‘On the Source of the Eternal Truths’, trans. Edward A. Beach, in Owl of Minerva 22(1) (1990): 55–67.

© Sebastian Gardner, 2016.

Sebastian Gardner is Professor of Philosophy at UCL in London, UK. His work concentrates on Kant and post-Kantian philosophy. He has published widely on Kant, Fichte, Schelling, Schopenhauer, Nietzsche, Sartre, Merleau-Ponty and the aesthetic turn in post-Kantianism. Among his books are Kant and the Critique of Pure Reason (Routledge 1999), Sartre’s “Being and Nothingness” (Continuum/Bloomsbury, 2009) and, together with Matthew Grist, The Transcendental Turn (Oxford UP 2015). He is currently working on a book on the legacy of Kant’s Third Critique.